Anthropic CCA-F (Claude Certified Architect - Foundations (CCAR-F)) Practice Test 19

Question 180

A document extraction job contains 8,000 unrelated forms. Which partitioning enables scalable throughput?

  1. Define parallel prerequisite branches with a join gate before completion
  2. Decouple read-only verification from the authorized state-changing step
  3. Begin by completing mandatory high-value deliverables before optional expansions
  4. Divide inputs into independent batches with stable document IDs

Explanation

Correct Answer

D:

Independent document chunks can run concurrently and recover separately. Unique IDs preserve attribution when outputs arrive out of order.

Incorrect Answers

A:

A fan-out and join pattern preserves concurrency while enforcing that no required verification branch is skipped; its benefits for joining mandatory parallel checks do not resolve the need for partitioning bulk documents into traceable units.

B:

Distinct stages expose eligibility decisions and permit human approval before mutations; supporting separating validation from side-effecting tasks cannot replace the required partitioning bulk documents into traceable units.

C:

A critical-path plan ensures that the required outcome exists even if time runs short; this contributes to prioritizing the critical reporting path without meeting the need for partitioning bulk documents into traceable units.

Question 181

A support assistant is uncertain whether a customer qualifies under an undocumented exception. Which behavior is safest?

  1. Abstain from automatic commitment and send the exception to the authorized owner
  2. Provide policy basis, supporting records, and the requested action before approval
  3. Rank cases by material impact, uncertainty and business risk
  4. Mandate an authenticated human approval before the mutation is dispatched

Explanation

Correct Answer

A:

Policy ambiguity is not resolved by fluent reasoning. A controlled exception channel lets a human decide without making an unauthorized promise.

Incorrect Answers

B:

Meaningful approval requires adequate context to judge eligibility and consequences; that may serve giving human reviewers sufficient decision evidence, but it is no substitute for routing uncertain policy exceptions to authorized review.

C:

Triage uses the cost of errors to allocate scarce review capacity; this may support directing human attention to the highest-risk cases but cannot alone meet the requirement for routing uncertain policy exceptions to authorized review.

D:

Approval must be checked at the controller boundary rather than inferred from model confidence; the stated problem requires routing uncertain policy exceptions to authorized review rather than enforcing human authorization for high-value actions.

Question 182

A finance MCP integration has confused refund currencies, and a separate customer lookup is misused outside its intended data scope. Which TWO tool contract improvements solve those distinct problems? Choose TWO.

  1. Describe each customer lookup tool's exact resource and permitted query boundaries
  2. Specify typed required parameters with separate identifier semantics
  3. Provide a stable output schema and version incompatible changes
  4. Specify numeric refund amounts with explicit units and a mandatory currency parameter

Explanation

Correct Answers

A:

The design must include exposing the resource scope of retrieval tools: effective tool descriptions explain when the operation should be used and what it returns; vague statements force the model to infer missing API boundaries.

D:

Financial tools need unambiguous units and typed fields. Encoding currency in the schema avoids silent assumptions when the same numeric value can represent different liabilities.

Incorrect Answers

B:

A precise input schema can distinguish account identifiers from email addresses and reject unsupported shapes; this addresses distinguishing identity fields in a tool interface, while the question is about disambiguating monetary units in tool calls.

C:

A predictable contract lets agents and validators consume results consistently; that relates to managing output schema compatibility, not the necessary disambiguating monetary units in tool calls.

Question 183

A cost-sensitive workflow keeps making redundant database calls after its evidence is sufficient. What control bounds that behavior?

  1. Set per-task call budgets and terminate or escalate on exhaustion
  2. Disambiguate each tool's selection criteria and record type in descriptions
  3. Refuse the unauthorized call and report a permission denial to orchestration
  4. Allow a privileged diagnostics component inspect logs and return sanitized findings

Explanation

Correct Answer

A:

Tool budgets prevent unbounded external work and let the system deliver a controlled incomplete status. Pure prompt advice is less predictable for hard usage limits.

Incorrect Answers

B:

Tool routing quality depends on distinct metadata; the unsatisfied requirement is controlling tool invocation cost per task, not disambiguating adjacent retrieval tools.

C:

A capability boundary must be enforced at execution time; in this scenario, enforcing a denied capability at runtime is secondary to controlling tool invocation cost per task.

D:

When considering segregating diagnostic access from regular agents, dedicated diagnostics isolates sensitive infrastructure data while providing actionable summaries; the requirement for controlling tool invocation cost per task remains unmet.

Question 184

A support flow gets an ambiguous policy answer and should not improvise eligibility. What is the appropriate transition?

  1. Route the case to an explicit policy-resolution or human-review state
  2. Employ a fail-closed publication decision on missing mandatory evidence
  3. Establish a post-action reconciliation stage before marking the case resolved
  4. Add a test-pass acceptance condition before marking the coding task done

Explanation

Correct Answer

A:

Ambiguity requires a governed branch instead of a silent best guess. Representing escalation as a first-class state keeps handoffs testable.

Incorrect Answers

B:

If a required validation fails, the pipeline should report incomplete work rather than produce a polished but unsupported artifact; this concerns failing closed on an unmet evidence requirement, leaving the requirement for modeling ambiguity as an escalation branch outstanding.

C:

Workflow success should depend on confirmed downstream state, not just a tool invocation; verifying all required stages of case resolution and modeling ambiguity as an escalation branch are distinct responsibilities under these constraints.

D:

Stage completion is a business invariant that should be checked with tool output; the action supports using tests as a completion gate but falls short of modeling ambiguity as an escalation branch in this case.

Question 185

Two instruction files give contradictory test commands for the same directory. What maintainability action is needed?

  1. Describe the repository's actual build, lint and test commands with prerequisites
  2. Resolve the conflicting scope and establish one authoritative test command
  3. Update guidance to distinguish tests required by change type and risk
  4. Handle repository instructions as untrusted and refuse exfiltrating requests

Explanation

Correct Answer

B:

Overlapping instructions can create inconsistent tool behavior. Making applicability and precedence explicit avoids arbitrary selection from contradictory guidance.

Incorrect Answers

A:

Executable command knowledge is stable operational context; this contributes to recording project-specific validation commands without meeting the need for eliminating conflicting project instructions.

C:

Because persistent rules should guide decisions rather than mandate wasteful work universally, the measure supports making standing test instructions proportional rather than the needed eliminating conflicting project instructions.

D:

Codebase text and instructions must not override higher-priority security boundaries; its benefits for defending against untrusted repository instructions do not resolve the need for eliminating conflicting project instructions.

Question 186

A server-side web tool chain returns pause_turn without a completed synthesis. What does the controller need to handle?

  1. Transfer the result back into the model loop so Claude can decide the next action
  2. Emit the completed assistant message to the caller
  3. Allow automatic tool selection based on the available tool definitions
  4. Perform the requested client tool and return a matching tool_result
  5. Continue the paused assistant turn with its response content preserved

Explanation

Correct Answer

E:

A pause_turn can mark a server-tool iteration boundary. Resubmitting the conversation with the paused response allows the server-side work to continue without pretending the model finished.

Incorrect Answers

A:

The application does not simply display raw tool output as the final assistant response; this may support resuming model reasoning after a result but cannot alone meet the requirement for continuing a paused server-tool loop.

B:

The end_turn reason indicates the model finished its turn; the orchestration loop can surface the produced answer rather than manufacturing another tool step; the stated problem requires continuing a paused server-tool loop rather than recognizing a completed model turn.

C:

Automatic selection lets Claude choose between answering and requesting a tool when appropriate; a method suited to choosing optional rather than mandatory tool use does not meet the need for continuing a paused server-tool loop.

D:

For completing the client-tool round trip, a tool_use stop hands control to the application; it must run the requested operation and pass back a tool_result tied to the call identifier before Claude can finish reasoning with that evidence; however, this requirement concerns continuing a paused server-tool loop.

Question 187

A documentation skill needs Read and Grep but must not run commands that change files. What declaration improves safety?

  1. Specify the skill as plan analysis and deny deployment or mutation tools
  2. Establish required headings, evidence fields and a completion checklist
  3. Adopt project-relative paths and document required environment inputs
  4. Restrict skill tool permissions to the necessary read-oriented operations
  5. Build a project skill containing the ADR template and review criteria

Explanation

Correct Answer

D:

A skill's allowed tool surface should match its declared job. Narrow capabilities lower the chance of accidental changes during a read-only documentation audit.

Incorrect Answers

A:

Skill language and access control should agree; that relates to separating review skills from deployment skills, not the necessary defining least-privilege tools for a skill.

B:

Although a bounded output contract makes a reusable workflow easier to verify, the choice addresses standardizing a command's output shape and not the requested defining least-privilege tools for a skill.

C:

Portable skills avoid machine-specific assumptions and resolve resources from a predictable project context; the deciding condition is defining least-privilege tools for a skill, whereas this action concerns removing machine-specific assumptions from a skill.

E:

A skill can provide consistent document structure and decision checkpoints; this addresses reusing a controlled architecture-record template, while the question is about defining least-privilege tools for a skill.

Question 188

A large repository contains generated source files that should not be edited manually. What discovery should precede a patch?

  1. Collect a short verified configuration summary with exact file references
  2. Determine references and consumer contracts before changing the shared definition
  3. Trace route registration and callers from the known endpoint name
  4. Identify generation sources and edit the generator or schema instead of generated output

Explanation

Correct Answer

D:

Generated artifacts are often overwritten in the next build. Tracing their source ensures the change survives regeneration and follows repository workflow.

Incorrect Answers

A:

Shared grounded notes prevent repeated expensive reads while retaining the ability to verify details; in this scenario, reusing concise verified repository findings is secondary to distinguishing generated files from authoritative sources.

B:

When considering mapping consumers of a shared interface, shared APIs can affect distant packages; the requirement for distinguishing generated files from authoritative sources remains unmet.

C:

When addressing tracing code paths from a known endpoint, a dependency-directed traversal identifies the local change impact; the present task instead calls for distinguishing generated files from authoritative sources.

Question 189

An architecture diagram shows every subagent calling every other worker directly without limits. Which change reduces coupling?

  1. Record each branch's completion state and retry only the failed dependency
  2. Launch independent read-only specialist tasks concurrently and join their outputs
  3. Allocate analysis-only capabilities to the reviewer and editing tools to a separate implementer
  4. Route cross-specialty requests through an orchestrator with an explicit task contract

Explanation

Correct Answer

D:

Hub-and-spoke mediation controls ownership, context, and failure propagation. Arbitrary direct calls create cycles and obscure which worker is responsible for the final result.

Incorrect Answers

A:

Durable per-worker status lets the orchestrator resume a specific failed task; recovering a failed worker without rerunning completed branches and minimizing unpredictable peer-to-peer dependencies are distinct responsibilities under these constraints.

B:

When branches have no data dependency, parallel execution reduces waiting time while leaving the coordinator responsible for combining results and handling any branch failure; the action supports parallelizing independent evidence gathering but falls short of minimizing unpredictable peer-to-peer dependencies in this case.

C:

Tool partitioning enforces role boundaries; this can serve separating creator and reviewer privileges, but leaves the requirement for minimizing unpredictable peer-to-peer dependencies unmet.

Question 190

A support response must avoid promising an action until the API confirms it. What should the instructions say?

  1. State required positive findings, limitations, risks and outstanding decisions
  2. Require tool-confirmed completion before wording any definitive customer promise
  3. Define explicit eligibility conditions and required evidence for each refund decision
  4. Assign decisions and proposals to named sources, preserving unresolved disagreement

Explanation

Correct Answer

B:

The prompt can distinguish proposed, attempted and confirmed actions. This prevents the assistant from inferring financial success from its own intent.

Incorrect Answers

A:

Because balanced section requirements make omissions more visible during validation, the measure supports eliciting both benefits and risks in reports rather than the needed preventing premature promises of side effects.

C:

Concrete criteria make a model's judgment testable against policy; supporting translating broad intent into checkable refund rules cannot replace the required preventing premature promises of side effects.

D:

Attribution is part of factual correctness; this contributes to preserving attribution in generated meeting summaries without meeting the need for preventing premature promises of side effects.

Claude Certified Architect - Foundations (CCAR-F) (CCA-F) Practice Tests